More universities are starting master’s programs in data science and analytics, of which statistics is foundational, due to the wide interest from students and employers. Amstat News reached out to those in the statistical community who are involved in such programs. Given their interdisciplinary nature, we identified programs involving faculty with expertise in different disciplines to jointly reply to tour questions. We profiled a few universities in our April and June issues; here are several more, plus a few PhD programs. [Harvard, University of British Columbia, Texas A&M, University of Colorado-Denver, University of Wisconsin, University of Central Florida]
A sport-specific data team at Canadian Tire has been working with summer and winter sports federations to gain the critical advantages in training and preparation
It looks like Spotify AB really is going to go public by doing a direct listing on a U.S. stock exchange, and I am starting to get excited. The idea of a direct listing is that, instead of doing an underwritten initial public offering in which sellers (Spotify and its founders and early investors) decide how much they want to sell, sign up some banks, build a book of demand, and then all at the same time sell their stock to investors chosen by the underwriters, Spotify will just one day declare that it is public and that anyone who wants to buy or sell can, on the stock exchange, like any other stock.
In practice I assume this means that Spotify will go public by means of an opening auction on the New York Stock Exchange or Nasdaq: Early one morning, some Spotify shareholders will make indicative offers to sell their shares, and some bold investors will make indicative bids to buy them, and the exchange will publish some tentative price that seems like it will clear the supply and demand, and then other shareholders and buyers can come in and adjust the prices and quantities that they want, and eventually a clearing price will be reached, and the stock will open and trade normally.
In a world that requires increasing amounts of compute power to handle the resource-intensive demands of workloads like artificial intelligence and machine learning, IBM enters the fray with its latest generation Power chip, the Power9.
The company intends to sell the chips to third-party manufacturers and to cloud vendors including Google. Meanwhile, it’s releasing a new computer powered by the Power9 chip, the AC922 and it intends to offer the chips in a service on the IBM cloud. “We generally take our technology to market as a complete solution,” Brad McCredie, IBM fellow and vice president of cognitive systems explained.
Nature, Comment; Phil De Luna, Jennifer Wei, Yoshua Bengio, Alán Aspuru-Guzik & Edward Sargent
from
Artificial intelligence can speed up research into new photovoltaic, battery and carbon-capture materials, argue Edward Sargent, Alán Aspuru-Guzikand colleagues.
Many firms today are introducing cognitive technologies to their organizations somewhat slowly. It’s not that they don’t believe the technologies are important, but rather that they have other, more pressing priorities, or that they need to prepare their environments for effective AI implementation. The Bank of Montreal is one organization that is moving steadily toward this objective. … For the last several years, BMO has initiated a series of transformations to its technology infrastructure under the leadership of Jean-Michel Arès, the Group Head of Technology and Operations, and François Joanette, the bank’s Chief Data Officer. Like many large banks, complying with regulatory requirements has been a top priority. The bank also needed to update its basic processes for storing and reporting on data. Data science and cognitive technologies were certainly of interest to the bank, but since they both rely heavily on large volumes of high-quality data, these new technologies needed to wait for the infrastructure improvements.
After being programmed with only the rules of chess (no strategies), in just four hours AlphaZero had mastered the game to the extent it was able to best the highest-rated chess-playing program Stockfish.
In a series of 100 games against Stockfish, AlphaZero won 25 games while playing as white (with first mover advantage), and picked up three games playing as black. The rest of the contests were draws, with Stockfish recording no wins and AlphaZero no losses.
“We now know who our new overlord is,” said chess researcher David Kramaley, the CEO of chess science website Chessable.
ARTIFICIAL intelligence (AI) has already changed some activities, including parts of finance like fraud prevention, but not yet fund management and stock-picking. That seems odd: machine learning, a subset of AI that excels at finding patterns and making predictions using reams of data, looks like an ideal tool for the business. Yet well-established “quant” hedge funds in London or New York are often sniffy about its potential. In San Francisco, however, where machine learning is so much part of the furniture the term features unexplained on roadside billboards, a cluster of upstart hedge funds has sprung up in order to exploit these techniques.
These new hedgies are modest enough to concede some of their competitors’ points. Babak Hodjat, co-founder of Sentient Technologies, an AI startup with a hedge-fund arm, says that, left to their own devices, machine-learning techniques are prone to “overfit”, ie, to finding peculiar patterns in the specific data they are trained on that do not hold up in the wider world. This is especially true of financial data, he says, because of their comparative paucity. Share-price time series going back decades still contain far less information than, say, the image data used to train Facebook’s facial-recognition algorithms.
As Spotify continues to inch towards a public listing, Apple is making a move of its own to step up its game in music services. Sources tell us that the company is close to acquiring Shazam, the popular app that lets people identify any song, TV show, film or advert in seconds, by listening to an audio clip or (in the case of, say, an ad) a visual fragment, and then takes you to content relevant to that search.
Randy Howard Katz, who helped develop many of the wireless tools and fast, reliable computer storage we take for granted today, has been appointed vice chancellor for research at UC Berkeley.
Asu Ozdaglar, the Joseph F. and Nancy P. Keithley Professor of Electrical Engineering and Computer Science, has been named the new head of the Department of Electrical Engineering and Computer Science (EECS), effective Jan. 1, 2018. She has been the interim head of the department since July 1, 2017, when former head Anantha Chandraksan was named dean of the School of Engineering.
“Professor Ozdaglar is an inspiring researcher and has emerged as a true leader in the areas of optimization theory and algorithms, game theory, and networks,” Chandrakasan says. “Her vision and dedication as an educator have been equally impressive. She is both a tireless advocate and coach for her students, and she has been a strong advocate for educational innovation in EECS.”
“The lunar mapping work the FDL 2017 Lunar Water and Volatiles team is doing is more relevant than ever” … “Can you identify more craters than our Lunar Crater Detector? Click Here to Play!”
Attendees at the Faculty-Meets-Faculty luncheon held on November 29 were treated to a presentation by Professor of Computer Science and Engineering Paul Torrens, who is also affiliated with NYU’s Center for Urban Science and Progress (CUSP). Torrens develops and applies computer-based modeling and simulation tools that can be used for a variety of purposes — including studying pedestrian mobility along streetscapes, planning evacuation strategies in the case of a building collapse or other such catastrophe, and developing ways to control autonomous vehicles using physical gestures, to name a few of the projects on which he is working. For example, while Ford has long used LIDAR (Light Detection and Ranging) in its pre-collision detection systems, the company is now funding Torrens’s research into physical movement and gestures, in order to prevent autonomous vehicles from hitting pedestrians.
Torrens explained that technologies like LIDAR can provide big data, but now, such data can be supplemented by what he characterized as “big awareness.”
NYU’s Claudio Silva & Juliana Freire, both Professors of Computer Science, Engineering and Data Science, develop ARIES (ARt Image Exploration Space) with a multidisciplinary team of researchers
By Mike Innes, David Barber, Tim Besard, James Bradbury, Valentin Churavy, Simon Danisch, Alan Edelman, Stefan Karpinski, Jon Malmaud, Jarrett Revels, Viral Shah, Pontus Stenetorp and Deniz Yuret
from
As programming languages (PL) people, we have watched with great interest as machine learning (ML) has exploded – and with it, the complexity of ML models and the frameworks people are using to build them. State-of-the-art models are increasingly programs, with support for programming constructs like loops and recursion, and this brings out many interesting issues in the tools we use to create them – that is, programming languages.
While machine learning does not yet have a dedicated language, several efforts are effectively creating hidden new languages underneath a Python API (like TensorFlow) while others are reusing Python as a modelling language (like PyTorch). We’d like to ask – are new ML-tailored languages required, and if so, why? More importantly, what might the ideal ML language of the future look like?
This is a “proposed Q&A site for researchers and practitioners who use computers to model, simulate, and analyze phenomena in social sciences and humanities, such as computational economics, computational sociology, cliodynamics, culturomics, and contents analysis.”